About Me
Hi, I'm Sebastian Tampu, a machine learning engineer and data scientist with a background in physics. I graduated with an MEng in computer engineering from the University of Toronto, with an emphasis in analytics, and machine learning. Before that, I completed my BSc in physics, which I still have a deep interest in.
I’m passionate about blending data and creativity through AI. Some of my key projects so far include clustering ETFs for financial strategies, creating a chatbot inspired by "Lord of the Rings", and developing a CNN-based emotion recognition model. I'm also exploring applications in fields I enjoy, like physics, strategy games, video games, and sport.
In my free time, I enjoy playing Go (the board game), world-building for D&D campaigns, and watching Borrussia Dortmund (never a moment of peace with them) and Formula 1. I’m also working on improving my language skills and fitness. These activities keep me curious and balanced outside of my professional life.
My goal is to create innovative AI solutions that leave a tangible impact across industries while continuously learning and growing. I’m particularly excited about the potential to combine AI with my personal interests to develop meaningful, creative, and cutting-edge projects. Take a look at my experiences and projects and feel free to reach out in any way you prefer!
Experience
Key roles I've held, demonstrating my expertise in data analysis, machine learning, and research:
Data Analyst
TELUS International | January 2023 – July 2024
Evaluated online search results for relevance and user satisfaction, improving algorithm accuracy.
Managed media content quality evaluations, setting metrics to optimize audio, image, and video search.
Enhanced webpage summarization processes to meet user information needs more effectively.
Data Analyst Intern
University of Toronto | September 2021 – April 2022
Analyzed semiconductor silicon test chips for the ATLAS detector at CERN.
Used Python for data analysis and visualization, identifying optimal testing temperatures.
Found key correlations between temperature and performance, aiding detector efficiency.
Data Scientist Intern
Toronto Metropolitan University | June 2020 – August 2020
Applied machine learning to low-dose CT denoising, improving image quality metrics.
Conducted a literature review on denoising structures, enhancing model architectures.
Built Python-based models for image denoising and performance analysis.
Projects
Some of my key projects, showcasing technical skills and problem-solving abilities:
ETF Clustering and Rotational Momentum Strategy Optimization
Applied hierarchical and k-means clustering to classify ETFs by returns and volatility. This project optimized a rotational momentum strategy, identifying high-performing ETFs for portfolio enhancement.
GitHubNano Language Model for LOTR-style Content & Chatbot
Created a custom transformer and fine-tuned DialoGPT in PyTorch to generate “Lord of the Rings”-inspired content and character interactions.
GitHubEmotion Recognition Model
Developed CNN-based models for emotion recognition, tuning AlexNet, ResNet, VGG16, and custom architectures for enhanced accuracy.
GitHubParallel Implementation on Jetson Nano
Implemented RRT and RRT* path planning algorithms in C++ and CUDA on a Jetson Nano, achieving near real-time performance through parallel execution.
No link available (Private Project)